Many trainees lack competence in performing cold snare polypectomy (CSP), and longer observation periods using assessment tools, such as the Cold Snare Polypectomy Assessment Tool (CSPAT), may be required. However, these tools are not commonly used in busy academic endoscopy practices. This study evaluates the concordance between trainee self-assessment of CSP with expert assessment and assesses factors associated with concordance. Consecutive CSPs performed by gastroenterology trainees from 2 institutions were video recorded and rated by 8 blinded experts from 4 different academic institutions in the United States and Singapore using the CSPAT from August 2017 to February 2020. Trainees self-assessed competence for each CSP immediately after the procedure. Concordance between trainee and trainer was reported as percentage of agreement in competence (score of 3 or 4) or not yet competent (score of 1 or 2). Twenty-two trainees performed 765 colonoscopies with 2,267 CSPs. Concordance was found between expert and trainee assessment of competence for 1,380 CSPs (60.9%; 95% CI, 58.8%-62.9%; weighted κ = 0.12; 95% CI, 0.08-0.16). Trainees underassessed competence for 541 CSPs (23.9%; 95% CI, 22.1%-25.7%) and overassessed competence for 326 CSPs (15.3%; 95% CI, 13.8%-16.8%). Career plan of private practice general gastroenterology (adjusted odds ratio, 0.40; 95% CI, 0.17-0.92) and lower colonoscopy volume before study (adjusted odds ratio, 0.35; 95% CI, 0.22-0.62) were independently associated with lower likelihood of concordance between expert and trainee assessment of competence. Structured feedback was not associated with concordance, and there was no change as trainees gained experience; however, feedback decreased the proportion of overassessments. There is poor concordance between trainee self-assessment and trainer assessment of CSP. Self-assessment can be used as an initial part of competency assessment, which should then be reconciled with external assessments to improve concordance in assessments.
Read full abstract